Hi SMY14INCH,
In my 40 or so years in audio service, the link between temperature and defects has been demonstrated practically time and time again. As well, 5 years is nothing. There has been high numbers of computers break down do to temperatures being too high. I recently had a modified Hafler DH-110 come in. The signal transistors ran far too hot. Possibly like your own preamp does. It needs a set of transistors as they are all suspect now with a few failing intermittently. High temperatures are the enemy to trouble-free performance in electronics. 3 months is nothing. I'm also working on 40 year old products. That is reliability.
-Chris
In my 40 or so years in audio service, the link between temperature and defects has been demonstrated practically time and time again. As well, 5 years is nothing. There has been high numbers of computers break down do to temperatures being too high. I recently had a modified Hafler DH-110 come in. The signal transistors ran far too hot. Possibly like your own preamp does. It needs a set of transistors as they are all suspect now with a few failing intermittently. High temperatures are the enemy to trouble-free performance in electronics. 3 months is nothing. I'm also working on 40 year old products. That is reliability.
Per transistor, or total?
Of course! This has zero to do with temperature and everything to do with making the change of current with the signal in each transistor a lower percentage. You also gain a lower output impedance doing this. These things are not affected by how hot the transistor is unless it dies. Try not to confuse current flow with temperature.
-Chris
Just by coincidence (I believe) in EDN:
A negative feedback model for transistors with emitter/source degeneration | EDN
Jan
A negative feedback model for transistors with emitter/source degeneration | EDN
Jan
As I said, I was merely showing where your logic leads. All circuits can be analysed using circuit theory alone. Many circuits (almost all circuits, as I have shown!) can also be analysed using feedback theory. Your argument was that as an emitter follower (at low frequencies - although you didn't say that) can be analysed using circuit theory then it cannot use feedback. This is logically false, because it assumes that any circuit can have one and only one analysis. I merely demonstrated that a circuit using just resistors can be analysed using feedback theory, thus (according to your logic) it must be analysed using feedback theory alone and not circuit theory.traderbam said:
The advantage of using feedback theory (where it is appropriate) is that it gives useful insights into circuit operation (e.g. stability) which would be harder to determine (although still possible) using pure circuit theory. Feedback theory is certainly applicable to an emitter follower (as I have demonstrated); it is appropriate to an emitter follower because it gives insights about stability and input impedance and distortion which might be harder to obtain from the naive 'potential divider' model. In any case, it is blindingly obvious (to almost everyone) that an EF subtracts the whole output signal from the input and therefore has 100% feedback.
What is slightly more subtle, and so catches out more people, is that emitter degeneration (with the output taken from the collector) is also feedback.
Actually I find that rather obvious, if you realize that the power supply rail is AC ground; connecting the top of Rc with the bottom of Re and voilá - a regular feedback divider ;-). It even follows the familar gain equation!
Only shows that any attempt to fool mother nature is doomed at its outset!
Jan
Interesting. My design goals are that varying supply doesn't affect biasing, and this is precisely to get the best performance possible by rejecting power supply changes. In fact, neither driver transistors nor VAS transistors need to be the same, let alone matched (although the same transistors should be used).
My latest design can be used with 12-0-12 transformer up to 18-0-18 without changing a single component. Biasing should remain constant. The only thing that changes is output power.
But then again, class A is a different art.
Just so you guys do not get confused I am talking about having fixed vcc so voltage gain stage transitor bias can stay in best optimum operating point for extra sound.
Yes, but why? Isn't your VAS controlled by a current source? I know you can use bootstrapping and other resistive current sources, but a current source in the VAS is the absolute bees knees.
This is one factor out of many that will make my amp diffrent and sound better than others.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.